KR101153011B1 - Designing method of photomask, manufacturing method of photomask and computer readable recording medium having design program of photomask - Google Patents

Abstract

According to an embodiment, a procedure for dividing design pattern data into predetermined regions, a procedure for obtaining a peripheral length of the pattern in the divided region, and a repeating of the above procedure for obtaining the peripheral length of the pattern in the entire design pattern data A procedure for obtaining a dimensional conversion difference in the whole design pattern data from the correlation between the procedure, the circumferential length of the pattern in the whole design pattern data, the circumferential length of the pattern previously determined, and the dimensional conversion difference; A process for performing a process conversion difference correction of design pattern data using the value of and a procedure for creating exposure pattern data from the corrected design pattern data are provided.

Description

DESIGNING METHOD OF PHOTOMASK, MANUFACTURING METHOD OF PHOTOMASK AND COMPUTER READABLE RECORDING MEDIUM HAVING DESIGN PROGRAM OF PHOTOMASK}

<Related application>

This application is based on Japanese Patent Application No. 2009-218226 for which it applied on September 21, 2009, and claims its priority, The whole content is integrated in this specification as a reference.

Embodiment of this invention relates to the design method of a photomask, the manufacturing method of a photomask, and the design program of a photomask.

In a recent semiconductor device and the like, the pattern is made finer, and as the design rule becomes smaller, it becomes difficult to transfer the design pattern onto the substrate (for example, a wafer) with high accuracy. For example, in the case of transferring a design pattern onto a wafer, when the difference between the design pattern and the pattern on the transferred wafer becomes large, electrical characteristics may deteriorate, bridges or disconnections of the pattern may occur, or the yield may decrease. There is a concern.

For this reason, process conversion difference correction (PPC: Process Proximity Correction) is applied to correct the shape of the mask pattern so as to be transferred according to the shape and dimensions of the design pattern.

Here, it is known that the deviation of the pattern dimension formed by the so-called microloading effect in an etching process becomes large.

Therefore, the manufacturing method of the photomask which performs the process conversion difference correction which considered the micro loading effect is proposed (refer patent document 1).

In the technique disclosed in Patent Document 1, the process area difference correction data for each pattern area ratio (covering rate) calculated in advance is calculated by calculating the pattern area ratio (covering rate) in the region including the pattern to be corrected. The design pattern is corrected based on the calculated pattern area ratio (coating rate).

Japanese Patent Laid-Open No. 2004-333529

However, in the technique disclosed in Patent Document 1, no consideration has been given to the density of the pattern. Therefore, when the process conversion difference correction data calculated | required previously are used, there exists a possibility that the error may arise in dimensional conversion difference prediction according to the density of a pattern. Moreover, about the pattern to which the refinement | miniaturization progresses in recent years, there existed a possibility that the error of the dimensional conversion difference prediction might become larger.

BRIEF DESCRIPTION OF THE DRAWINGS The schematic plan view for demonstrating the case where a dimension conversion difference differs even if "the pattern area ratio (coating rate)" is the same.
2 is a schematic diagram for illustrating the relationship between the dimensional conversion difference and the surface area of the pattern side surface.
3 is a schematic diagram illustrating a method of distinguishing a dense pattern region and a sparse pattern region.
4 is a schematic diagram illustrating another method of distinguishing a dense pattern region and a sparse pattern region.
5 is a schematic diagram for illustrating a case of predicting a dimensional conversion difference in consideration of gas flow (diffusion).
6 is a flowchart for illustrating a method of designing a photomask according to the present embodiment.

According to an embodiment, a design is repeated by repeating a procedure for dividing design pattern data into predetermined regions, a procedure for obtaining a peripheral length of the pattern in the divided region, and a procedure for calculating a peripheral length of the pattern in the divided region. A step between obtaining a peripheral length of the pattern in the entire pattern data, a peripheral length of the pattern determined in a procedure for obtaining a peripheral length of the pattern in the design pattern data, and a preliminary difference between the peripheral length of the pattern and the dimension conversion difference From the obtained correlation, the procedure for calculating the dimensional conversion difference in the whole design pattern data, the process of correcting the process conversion difference of the design pattern data using the value of the calculated dimensional conversion difference, and the design pattern data after the correction And a procedure for creating exposure pattern data from the same. A method of designing a photomask is provided.

According to another embodiment, there is provided a photomask manufacturing method comprising creating exposure pattern data using the above-described photomask design method and creating a photomask based on the created exposure pattern data. .

Further, according to another embodiment, a computer program includes a procedure of dividing design pattern data into a predetermined region, a procedure of calculating a peripheral length of the pattern in the divided region, and a peripheral length of the pattern in the divided region. By repeating the procedure of calculating, the peripheral length of the pattern obtained in the procedure of calculating the peripheral length of the pattern in the whole design pattern data, the procedure of calculating the peripheral length of the pattern in the whole design pattern data, and The process conversion difference of the design pattern data using the procedure of calculating the dimensional conversion difference in the whole design pattern data from the previously obtained correlation between the peripheral length of the pattern and the dimensional conversion difference, and the value of the obtained dimensional conversion difference. Exposure pattern data from the procedure for performing correction and the design pattern data after the correction That is to execute a procedure to write a program recording the design of the photomask, characterized in the computer readable medium is provided.

EMBODIMENT OF THE INVENTION Hereinafter, embodiment is described, referring drawings.

In order to transfer to the size and shape of a design pattern, it is necessary to predict the dimension conversion difference in advance (dimension conversion difference prediction) and correct the pattern shape on the photomask (reticle) used in the lithography step. Therefore, if the precision of a dimension conversion difference prediction can be improved, a desired pattern can be formed with high precision.

Here, by specifying an analysis index relating to matters affecting the precision of the dimensional conversion difference prediction, and revealing a correlation between the analysis index and the dimensional conversion difference, the accuracy of the dimensional conversion difference prediction can be greatly improved.

Therefore, first of all, the experience gained by the present inventor is illustrated with respect to an analysis index relating to matters affecting the precision of the dimensional conversion difference prediction.

As an analysis index regarding matters affecting the accuracy of the dimensional conversion difference prediction, “pattern area ratio (coating rate)” can be exemplified. "Pattern area ratio (coating rate)" is the area of the pattern to occupy per unit area.

Here, when the "pattern area ratio (coating rate)" changes, the incident amount of the incident material (for example, radicals, ions, etc.) also changes, so that the dimensional conversion difference also changes accordingly.

Therefore, when a correlation between the "pattern area ratio (covering rate)" and the dimensional conversion difference is obtained in advance by experiment, simulation, or the like, the dimensional conversion difference can be predicted. That is, the area of the pattern which occupies per unit area can be calculated | required based on design pattern data, and a dimensional conversion difference can be predicted from the previously calculated correlation between "pattern area ratio (covering rate)" and a dimension conversion difference.

However, according to the experience obtained by the present inventors, even if the "pattern area ratio (coating rate)" is the same, the dimensional conversion difference may be different.

FIG. 1: is a schematic plan view for demonstrating the case where a dimension conversion difference differs even if "the pattern area ratio (coating rate)" is the same.

2 is a schematic diagram for illustrating the relationship between the dimensional conversion difference and the surface area of the side surface of the pattern. In the case shown in FIG. 1A, the "pattern area ratio (coating rate)" with respect to the pattern 100a is 50%. On the other hand, also in the case shown in FIG. 1B, the "pattern area ratio (coating rate)" with respect to the pattern 100b is 50%.

However, if the height dimensions (dimensions in the direction substantially perpendicular to the planar shape shown) of the pattern 100a and the pattern 100b are the same, the sum of the surface areas of the side surfaces 102b of the pattern 100b is equal to that of the pattern 100a. It is larger than the sum of the surface areas of the side surfaces 102a. In this case, the larger the surface area of the side surface, the larger the incident amount of the incident material (for example, radicals or ions). Therefore, the value of the dimension conversion difference in the same pattern shape 100c becomes large, so that surface area becomes large, as shown in FIG.

That is, even if the "pattern area ratio (coating rate)" is the same, the dimensional conversion difference is different. Therefore, when a dimension conversion difference is calculated | required based on "pattern area ratio (coating rate), an error will arise in dimensional conversion difference prediction. In addition, there is a fear that the error of the dimensional conversion difference prediction becomes larger for the pattern in which the miniaturization is recently performed.

Therefore, in this embodiment, the difference in the surface area of the side surface of a pattern is considered by making "perimeter length of a pattern" into an analysis index.

In this case, "the peripheral length of a pattern" can be made into the outer peripheral length in the planar shape of the pattern which occupies per unit area.

As mentioned above, when the "peripheral length of a pattern" becomes long, the surface area of the side surface of the pattern increases, so that the incident amount of the incident material (for example, radicals or ions) increases. Therefore, the dimensional conversion difference also increases accordingly.

Therefore, when the correlation between the "perimeter length of the pattern" and the dimensional conversion difference is calculated in advance by experiment, simulation, or the like, the dimensional conversion difference can be predicted. That is, the outer periphery length (the "perimeter length of the pattern") in the planar shape of the pattern occupied per unit area is calculated based on the design pattern data, and the dimension conversion difference is determined from the previously obtained correlation between the "perimeter length of the pattern" and the dimension conversion difference. Can be predicted.

By doing in this way, since the dimension conversion difference which reflects the difference of the surface area of the side surface of a pattern can be predicted, the precision of a dimension conversion difference prediction can be improved.

In this case, when the treatment is a film forming process or the like, the influence of the "pattern area ratio (coating rate)" becomes small, so that the dimensional conversion difference can be predicted using only the "peripheral length of the pattern". On the other hand, when the process is an etching process or the like, the influence of the "pattern area ratio (covering rate)" is increased, so that not only the "peripheral length of the pattern" but also the "pattern area ratio (covering rate)" is used to predict the dimensional conversion difference. It is preferable. In addition, even when the process is a film forming process or the like, it is possible to predict the dimensional conversion difference by using "ambient length of the pattern" and "pattern area ratio (coating rate)".

When predicting the dimensional conversion difference using the "perimeter length of the pattern" and the "pattern area ratio (covering rate)", the "perimeter length of the pattern" and "pattern area ratio (covering rate)" according to the density of the pattern. May be used separately, or may be used in combination of "a peripheral length of a pattern" and "pattern area ratio (coating rate)."

For example, since the influence of the "peripheral length of a pattern" becomes small in a sparse pattern area, a dimensional conversion difference can be predicted using "pattern area ratio (coating rate)." Moreover, in the case of a dense pattern area or a refined pattern, the influence of the "perimeter length of a pattern" becomes large, and it is possible to predict the dimensional conversion difference using the "perimeter length of a pattern". In this case, when the combination of the "perimeter length of the pattern" and the "pattern area ratio (coating rate)" is used, the accuracy of the dimensional conversion difference prediction can be further improved.

Here, when using the "circumferential length of a pattern" and "pattern area ratio (coating rate)" according to the density of a pattern, it is necessary to distinguish a dense pattern area and a sparse pattern area in a design pattern.

3 is a schematic diagram for illustrating a method of distinguishing a dense pattern region and a sparse pattern region.

When distinguishing the dense pattern region from the sparse pattern region, first, the design pattern data is divided into predetermined regions (for example, micro regions), and patterns included in the divided regions are extracted as shown in FIG. 3A. .

Next, as shown in FIG. 3B, the "peripheral length of the pattern" in the extracted pattern is obtained. Then, by dividing the obtained "peripheral length of the pattern" using a predetermined threshold value, the distinction between the dense pattern region (the portion where the "peripheral length of the pattern" is long) and the sparse pattern region (the portion where the "peripheral length of the pattern" is short) is distinguished. Is done.

Then, as shown in FIG. 3C, the dense pattern region is extracted. In addition, a sparse pattern region may be extracted.

By repeating the above-described procedure, the dense pattern region and the sparse pattern region in the entire design pattern data are distinguished.

4 is a schematic diagram illustrating another method of distinguishing a dense pattern region and a sparse pattern region.

First, the design pattern data is divided into predetermined regions (for example, small regions), and patterns included in the divided regions are extracted as shown in FIG. 4A.

Then, as shown in Fig. 4B, each pattern is extended in the width direction. In this case, it expands so that a pattern may contact each other in the part with the dimension between patterns short.

Next, as shown in FIG. 4C, a so-called merge process is performed. At this time, in the part where the patterns contact each other, the patterns are integrated by overlapping each other.

Next, as shown in FIG. 4D, the pattern subjected to the merge process is reduced in the width direction. At this time, it is reduced as much as it is expanded in FIG.

Next, as shown in FIG. 4E, the dense pattern region and the sparse pattern region are distinguished by an exclusive OR (XOR) between the pattern in FIG. 4A and the pattern in FIG. 4D.

As another analysis index regarding the matter which affects the precision of the dimensional conversion difference prediction, the "thing about spreadability" can be illustrated.

For example, in an etching process, the substance removed by the incident material (for example, radicals, ions, etc.) is spread | diffused and is discharged | emitted to a process space. In addition, in the film forming process, the incident material is diffused to reach the surface to be processed. Therefore, the dimensional conversion difference also changes depending on the diffusivity.

Here, since processes, such as formation of a pattern, are generally performed in the reduced pressure environment of several Pa about, the diffusibility can be analyzed in consideration of the flow of gas. In this case, when it is difficult to derive the diffusion equation, an approximate solution may be obtained by the Monte Carlo method or the like.

If the correlation between the gas flow (diffusion) and the dimensional conversion difference is obtained in advance by experiment or simulation, the dimensional conversion difference can be predicted. For example, the flow of gas (diffusion) is analyzed by analyzing the flow of gas (diffusion) based on design pattern data (for example, the density of patterns) and process data (for example, processing pressure, etc.). ) And the dimension conversion difference can be predicted from the previously obtained correlation between That is, "the thing about the diffusivity" which is an analysis index can be related to the correlation of the flow of the gas analyzed and the dimensional conversion difference based on the data extracted from the design pattern data, and process data.

In this case, a combination of "analysis of diffusivity (correlation between gas flow and dimensional conversion difference)" as an analysis index, "ambient length of pattern", and "pattern area ratio (coating rate)" It is also possible to predict the dimensional conversion difference. That is, the dimensional conversion difference may be predicted in consideration of the gas flow (diffusion).

5 is a schematic diagram for illustrating the case where the dimensional conversion difference is predicted in consideration of the gas flow (diffusion). 5A is a schematic diagram for illustrating the shape of a portion to be processed (for example, a portion to be removed by an etching process), and FIG. 5B is for illustrating the case where there is no gas flow (diffusion). It is a schematic diagram, and FIG. 5C is a schematic diagram for illustrating the case where gas flow (diffusion) is considered.

As shown in Fig. 5A, the vicinity of the to-be-processed portion 101 is partitioned into a lattice shape, and the flow (diffusion) of gas in each section is determined by simulation or the like.

In the case where there is no gas flow (diffusion), as shown in FIG. 5B, the treatment (for example, removal by etching) is performed only in the portion to be processed 101. In this case, the degree of processing is set to "1" in order to compare with the case where gas flow (diffusion) is considered.

On the other hand, in the case where gas flow (diffusion) is taken into consideration, processing (for example, removal by etching) is also performed around the portion to be processed 101 as shown in Fig. 5C. In this case, the degree of processing in the portion to be processed 101 is lower than in the case where there is no gas flow (diffusion) shown in FIG. 5B. Further, the further away from the portion 101 to be processed, the lower the degree of processing. In addition, in the case of the thing shown in FIG. 5C, as an example, the grade of the process in the to-be-processed part 101 is set to "0.8", and the grade of the process in the periphery of the to-be-processed part 101 is set to "0.2". .

In this case, the higher the degree of processing, the larger the dimensional change and the larger the dimensional conversion difference. Therefore, the relationship between the magnitude | size and shape of the to-be-processed part 101 as illustrated in FIG. 5C, and the grade of a process, that is, "a periphery length of a pattern", "pattern area ratio (coating rate)", and "diffusion property" Can be predicted from the relationship between the relationship between the relationship between the flow of the gas and the relationship between the flow of the gas and the dimensional conversion difference.

In addition, the flow of gas (diffusion) is affected by the density of the pattern. In general, gas flow (diffusion) tends to occur in the sparse pattern region, and gas flow (diffusion) tends to stagnate in the dense pattern region and the finer pattern.

As described above, the "perimeter length of the pattern" and the "pattern area ratio (covering rate)" are used separately according to the density of the pattern, or the "perimeter length of the pattern" and the "pattern area ratio (coating rate)" are used. It can be used in combination.

Therefore, according to the density of the pattern, the combination of "dispersion (related to gas flow and dimensional conversion difference)", "ambient length of the pattern", and "pattern area ratio (covering rate)" are combined. In this case, the correlation with the dimensional conversion difference in each case may be determined by experiment, simulation, or the like.

As exemplified above, the combination of "dispersion (related to gas flow and dimensional conversion difference)", "circumferential length of the pattern" and "pattern area ratio (coating rate)", A more accurate comprehensive analysis index can be obtained. And by using such an analysis index, the precision of a dimension conversion difference prediction can be improved further.

In the design method of the photomask which concerns on this embodiment, the dimension conversion difference is calculated | required based on the analysis index illustrated above, and the pattern shape of the photomask used in a lithography process is correct | amended using the value of the calculated dimension conversion difference.

Hereinafter, the design method of the photomask which concerns on this embodiment is illustrated.

6 is a flowchart for illustrating a method of designing a photomask according to the present embodiment.

First, design pattern data (data of a pattern formed on a substrate (for example, a wafer)) is created (step S1).

Next, the design pattern data is divided into predetermined areas (for example, micro areas) (step S2).

Next, the "peripheral length of the pattern" and the "pattern area ratio (coating rate)" in the divided area are obtained (step S3).

For example, the pattern contained in the divided area | region is extracted, and "a peripheral length of a pattern" and a "pattern area ratio (coating rate)" are calculated | required based on the extracted pattern.

Subsequently, by repeating the above-described procedure, the "perimeter length of the pattern" and the "pattern area ratio (coating rate)" in the entire design pattern data are obtained (step S4).

Next, a dimension conversion difference is calculated | required based on "a peripheral length of a pattern" and "pattern area ratio (coating rate)" (dimension conversion difference prediction is performed) (step S5).

For example, design pattern data is globally calculated from the correlation obtained in advance between the "perimeter length of a pattern", "pattern area ratio", and the "circumference length of a pattern", "pattern area ratio", and the dimension conversion difference in the design pattern data area. Find the dimensional conversion difference at.

In this case, when using the "perimeter length of a pattern" and "pattern area ratio (cover rate)" as an analysis index, the "perimeter length of a pattern" and "pattern area rate (cover rate)" calculated | required in step S4, The dimension conversion difference is calculated | required from the correlation calculated | required previously between "a peripheral length of a pattern", "pattern area ratio (covering rate)", and a dimension conversion difference.

When the dimensional conversion difference is predicted using the "perimeter length of the pattern" and the "pattern area ratio (covering rate)", the "perimeter length of the pattern" and the "pattern area ratio (coverage rate)" according to the density of the pattern. "Can be used separately, or it can use combining" a perimeter of a pattern "and" pattern area ratio (coating rate). "

In addition, the dimensional conversion difference is determined by combining the "diffusion property (related to the correlation between the flow of gas and the dimensional conversion difference)" and the "perimeter length of the pattern" and the "pattern area ratio (covering rate)". You may. That is, in the procedure for calculating the dimensional conversion difference, the correlation between the flow of the gas analyzed and the dimensional conversion difference analyzed based on the data extracted from the design pattern data and the process data may be taken into consideration.

Next, the process conversion difference correction of the design pattern data is performed using the value of the obtained dimension conversion difference (step S6).

Next, exposure pattern data is created from the design pattern data after correction (step S7).

Further, the corrected design pattern data may further include a procedure for verifying a portion that does not satisfy the design rule.

If the design pattern data after correction has a portion (for example, a risk point) that does not satisfy the design rule, the design pattern data is corrected, and the dimensional conversion difference is obtained from the corrected data. The value of can be used to redo the process conversion difference correction.

In addition, the optical proximity effect correction can be performed simultaneously in the procedure of performing the process conversion difference correction.

When performing optical proximity effect correction simultaneously, the test mask which changed the "perimeter length of a pattern" and the "pattern area ratio (coating rate)" is created first (step S11).

For example, a test mask having a desired "perimeter length of a pattern" and a "pattern area ratio (coating rate)" is created by spreading a dot pattern or the like in an area of a predetermined size and forming an opening.

Subsequently, transfer is performed using a test mask, and the correlation between the "perimeter length of the pattern", "pattern area ratio (coating rate)" and values of optical proximity effect correction is determined based on the measured values of the transferred pattern. (Step S12).

The correlation between the obtained "peripheral length of the pattern", "pattern area ratio (coating rate)" and the value of optical proximity effect correction is used when performing optical proximity effect correction in step S5. In addition, since a known technique can be applied about optical proximity effect correction, the description is abbreviate | omitted.

As described above, the photomask is designed.

In addition, creation of design pattern data is not necessarily required, and the created design pattern data may be used.

In addition, although the dimensional conversion difference prediction is performed based on the "peripheral length of a pattern" and the "pattern area ratio (covering rate)", the dimensional conversion difference prediction can also be made based on the "peripheral length of a pattern". have. However, when the dimensional conversion difference prediction is performed based on the "peripheral length of the pattern" and the "pattern area ratio (coating rate)", the prediction accuracy can be improved.

According to the design method of the photomask which concerns on this embodiment, since the dimension conversion difference can be calculated | required based on "a peripheral length of a pattern" and "pattern area ratio (covering rate), the precision of a dimension conversion difference prediction can be improved. have. Therefore, accurate process conversion difference correction can be performed.

In addition, depending on the density of the pattern, the "circumferential length of the pattern" and the "pattern area ratio (covering rate)" are used separately, or the combination of the "perimeter length of the pattern" and "pattern area ratio (coating rate)" is used. Therefore, the precision of the dimension conversion difference prediction can be further improved.

In addition, a dimensional conversion difference can be determined by appropriately combining "related to diffusivity (correlation between gas flow and dimensional conversion difference)", "ambient length of the pattern" and "pattern area ratio (covering rate)". By doing so, the accuracy of the dimension conversion difference prediction can be further improved.

Moreover, since the part which does not satisfy a design rule (for example, a danger point etc.) can be extracted correctly, the verification precision of design pattern data can also be improved.

Next, the manufacturing method of the photomask which concerns on this embodiment is illustrated.

In the manufacturing method of the photomask which concerns on this embodiment, exposure pattern data is created from design pattern data by the design method of the photomask mentioned above, and a photomask is created based on the created exposure pattern data. In this case, the photomask can be prepared using the etching method.

According to the manufacturing method of the photomask which concerns on this embodiment, since the dimension conversion difference can be calculated | required based on the above-mentioned "peripheral length of a pattern", "pattern area ratio (coating rate), etc., the precision of a dimension conversion difference prediction is improved. You can. Therefore, accurate process conversion difference correction can be performed, and a photomask with few dimension conversion differences can be obtained. In addition, extraction of portions (for example, danger points, etc.) that do not satisfy the design rule can also be performed accurately. As a result, a photomask with excellent product yield can be obtained.

Next, the manufacturing method of the electronic component which concerns on this embodiment is illustrated.

In addition, the manufacturing method of a semiconductor device is mentioned as an example and demonstrated.

The manufacturing method of a semiconductor device is a process of forming a pattern on a wafer by film formation, resist coating, exposure, development, etching, resist removal, etc., an inspection process, a cleaning process, a heat treatment process, an impurity introduction process, a diffusion process, a planarization process. It is implemented by repeating a some process, such as these. In the manufacturing method of such a semiconductor device, a photomask is manufactured by the manufacturing method of the photomask mentioned above, and exposure is performed using the photomask manufactured in that way.

In addition, since the technique of each known process is applicable to the thing other than the manufacturing method of the photomask mentioned above, those description is abbreviate | omitted.

In addition, although the manufacturing method of the semiconductor device was mentioned as an example and demonstrated as an example of the manufacturing method of the electronic component which concerns on this embodiment, it is not limited to this. For example, it can be widely applied to the manufacture of electronic components using photolithography techniques, such as the formation of a pattern in the manufacture of a flat panel display (for example, the formation of a pattern in a liquid crystal color filter or an array substrate). have.

According to the manufacturing method of the electronic component which concerns on this embodiment, a pattern can be formed using the photomask with few dimensional conversion differences. In addition, a pattern can be formed using a photomask in which extraction of a portion (for example, a danger point, etc.) that does not meet the design rule is performed correctly. Therefore, deterioration of the electrical characteristics due to deformation of the pattern, bridge or disconnection of the pattern can be suppressed, and the improvement of product yield, quality, etc. can be aimed at.

Next, the design program of the photomask which concerns on this embodiment is illustrated.

The design program of the photomask according to the present embodiment computes design pattern data on a computer, calculates a dimensional conversion difference based on at least the "peripheral length of the pattern", and designs using the calculated value of the dimensional conversion difference. Process conversion difference correction of the pattern data is executed, and the exposure pattern data is calculated from the corrected design pattern data. It is also possible to perform optical proximity effect correction at the time of process conversion difference correction. In addition, a dimensional conversion difference can be determined by appropriately combining "circumferential length of a pattern", "pattern area ratio (covering rate)", and "dispersibility (related to gas flow and dimensional conversion difference)". You can also do that.

In order to execute the design program of the photomask, the design program of the photomask according to the present embodiment is stored in a storage unit (not shown) provided in the computer. In this case, the design program of the photomask can be supplied to the computer in a state of being stored in a recording medium (not shown), for example, and stored in a storage unit (not shown) installed in the computer by reading. It is also possible to be stored in a storage unit (not shown) installed in the computer via a communication line or the like connected to the computer.

The following procedures (1) to (7) can be executed, for example, by a design program of a photomask stored in a storage unit (not shown) installed in a computer.

(1) A procedure for calculating design pattern data based on input from a database (not shown) or input by an operator.

(2) A procedure of dividing design pattern data into predetermined regions (for example, minute regions).

(3) A procedure for calculating "a peripheral length of a pattern" and a "pattern area ratio (covering rate)" in the divided area.

(4) A procedure for calculating "ambient length of a pattern" and "pattern area ratio (covering rate)" in the entire design pattern data by repeating the above-described procedure.

(5) A procedure for calculating the dimensional conversion difference based on the "peripheral length of the pattern" and "pattern area ratio (covering rate)".

(6) A procedure for performing process conversion difference correction of design pattern data using the calculated value of the dimension conversion difference.

(7) A procedure of creating exposure pattern data from design pattern data after correction.

In addition, in (6), optical proximity effect correction can also be performed at the time of process conversion difference correction.

In addition, a procedure for calculating design pattern data is not necessarily required, and the calculated design pattern data may be provided or extracted.

In addition, the design program of the photomask which concerns on this embodiment may be executed by a single calculating part, and may be distributed and executed by a some calculating part.

In addition, in the corrected design pattern data, a procedure for performing verification of a portion that does not satisfy the design rule may be provided.

And when there is a part (for example, danger point etc.) which did not satisfy a design rule in the design pattern data after correction | amendment, an operator can be made to notify. Further, the design pattern data is corrected for the part that does not meet the design rule, the dimension conversion difference is calculated on the data after the correction, and the process conversion difference correction is executed again using the calculated value of the dimension conversion difference. You may.

In addition, based on "perimeter length of a pattern", "pattern area ratio (covering rate)", "dispersibility (related to the correlation between gas flow and dimensional conversion difference)" and an analysis index, the dimensional conversion difference For the purpose of obtaining, the description thereof is omitted as it is the same as described above.

According to the design program of the photomask which concerns on this embodiment, since the dimension conversion difference can be calculated | required based on "a periphery length of a pattern" and "pattern area ratio (covering rate), the precision of a dimension conversion difference prediction can be improved. have. Therefore, accurate process conversion difference correction can be performed.

In addition, depending on the density of the pattern, the "circumferential length of the pattern" and the "pattern area ratio (covering rate)" are used separately, or the combination of the "perimeter length of the pattern" and "pattern area ratio (coating rate)" is used. Therefore, the precision of the dimension conversion difference prediction can be further improved.

In addition, a dimensional conversion difference is calculated by appropriately combining "related to diffusivity (correlation between gas flow and dimensional conversion difference)", "ambient length of the pattern" and "pattern area ratio (covering rate)". In this case, the precision of the dimensional conversion difference prediction can be further improved.

Moreover, since the part which does not satisfy a design rule (for example, a danger point etc.) can be extracted correctly, the verification precision of design pattern data can also be improved.

In the above, this embodiment was illustrated. However, the present invention is not limited to these techniques.

As for the embodiment described above, design changes made by those skilled in the art as appropriate are included in the scope of the present invention as long as the features of the present invention are provided.

For example, although the case regarding etching is illustrated, the case where film-forming is performed to a to-be-processed part can be similarly carried out.

In addition, each element with which each embodiment mentioned above can be combined as much as possible, and combining these elements is also included in the scope of the present invention as long as it contains the characteristics of this invention.

Claims (20)

As a design method of the photomask,
A procedure of dividing the design pattern data into predetermined areas,
A procedure for obtaining a peripheral length of the pattern in the divided region;
A procedure for obtaining the peripheral length of the pattern in the entire design pattern data by repeating the procedure for calculating the peripheral length of the pattern in the divided area;
Dimensional conversion difference in the whole design pattern data from the peripheral length of the said pattern calculated | required by the procedure which calculates the peripheral length of the said pattern in the whole design pattern data, and the previously calculated correlation between the peripheral length of the said pattern and the dimension conversion difference. And the procedure to find
A process of correcting the process conversion difference of the design pattern data using the obtained value of the dimensional conversion difference;
And a procedure for creating exposure pattern data from the design pattern data after the correction. The method of claim 1,
In the procedure for obtaining the periphery length of the pattern in the divided region, the pattern area ratio is further obtained,
In the procedure for obtaining the periphery length of the pattern in the whole design pattern data, the pattern area ratio in the whole design pattern data is further obtained,
In the procedure for obtaining the dimensional conversion difference, the design is obtained from a preliminary correlation between the circumferential length and the pattern area ratio of the pattern and the circumferential length and the pattern area ratio of the pattern and the dimensional conversion difference throughout the design pattern data. And the dimensional conversion difference in the entire pattern data is obtained. The method of claim 1,
In the procedure for obtaining the periphery length of the pattern in the divided region, the pattern area ratio is further obtained,
In the procedure for obtaining the periphery length of the pattern in the whole design pattern data, the pattern area ratio in the whole design pattern data is further obtained,
In the procedure for obtaining the dimensional conversion difference, at least one of the perimeter length of the pattern and the pattern area ratio in the entire design pattern data is selected according to the density of the pattern. . The method of claim 3,
In the procedure for obtaining the dimensional conversion difference, the pattern area ratio is selected in the sparse pattern region, and the peripheral length of the pattern is selected in the case of a dense pattern region or a fine pattern. The method of claim 3,
In the procedure for determining the dimensional conversion difference, the peripheral length of the pattern is selected in the case of the photomask used for the film forming process, and the pattern area ratio and the peripheral length of the pattern is determined in the case of the photomask used in the etching process. The method of designing a photomask, characterized in that selected. The method of claim 1,
In the procedure for obtaining the dimensional conversion difference, the correlation between the flow of gas and the dimensional conversion difference analyzed based on the data extracted from the design pattern data and the process data is considered. The method of claim 1,
And a procedure for performing verification of a portion of the design pattern data after the correction that does not satisfy the design rule. The method of claim 7, wherein
In the procedure of verifying the part which does not satisfy the said design rule, when the part which does not satisfy the design rule is detected, the said design pattern data is corrected, The design method of the photomask characterized by the above-mentioned. The method of claim 1,
In the procedure for performing the process conversion difference correction, optical proximity effect correction is performed simultaneously. The method of claim 6,
And the optical proximity effect correction is performed based on a previously obtained correlation between at least one of the surrounding length and the pattern area ratio of the pattern and the value of the optical proximity effect correction. An exposure pattern data is created using the design method of the photomask of Claim 1, and a photomask is created based on the created said exposure pattern data, The manufacturing method of the photomask characterized by the above-mentioned. A computer-readable recording medium recording a design program of a photomask,
On the computer,
A procedure of dividing the design pattern data into predetermined areas,
Calculating a peripheral length of the pattern in the divided region;
Repeating the procedure for calculating the peripheral length of the pattern in the divided region, thereby calculating the peripheral length of the pattern in the entire design pattern data;
Dimension transformation in the whole design pattern data from the peripheral length of the said pattern calculated | required in the procedure which calculates the perimeter length of the said pattern in the whole design pattern data, and the previously calculated correlation between the peripheral length of the said pattern and the dimension conversion difference. To calculate the difference,
A process of performing process conversion difference correction of the design pattern data using the obtained value of the dimension conversion difference,
A computer-readable recording medium having recorded thereon a design program of a photomask, wherein a procedure for creating exposure pattern data from the corrected design pattern data is executed. The method of claim 12,
In the procedure for calculating the peripheral length of the pattern in the divided region, the pattern area ratio is further calculated,
In the procedure for calculating the periphery length of the pattern throughout the design pattern data, further calculating the pattern area ratio throughout the design pattern data;
In the procedure for calculating the dimensional conversion difference, the lateral conversion length and the pattern area ratio of the pattern in the entire design pattern data, and the predetermined relationship between the circumferential length and pattern area ratio and the dimensional conversion difference of the pattern And computing the dimensional conversion difference across the entire design pattern data. The method of claim 12,
In the procedure for calculating the peripheral length of the pattern in the divided region, the pattern area ratio is further calculated,
In the procedure for calculating the periphery length of the pattern throughout the design pattern data, further calculating the pattern area ratio throughout the design pattern data;
In the procedure for calculating the dimensional conversion difference, at least one of the peripheral length of the pattern and the pattern area ratio in the entire design pattern data is selected according to the density of the pattern. A computer readable recording medium having recorded a program. The method of claim 14,
In the procedure for calculating the dimensional conversion difference, the pattern area ratio is selected in the sparse pattern region, and the peripheral length of the pattern is selected in the case of a dense pattern region or a fine pattern. Computer-readable recording medium having recorded thereon. The method of claim 14,
In the procedure for calculating the dimensional conversion difference, the peripheral length of the pattern is selected in the case of the photomask used for the film forming process, and the pattern area ratio and the peripheral length of the pattern in the case of the photomask used in the etching process. A computer-readable recording medium having recorded thereon a program for designing a photomask. The method of claim 12,
In the procedure for calculating the dimensional conversion difference, the correlation between the flow of the gas and the dimensional conversion difference analyzed based on the data extracted from the design pattern data and the process data is considered. Recorded computer-readable recording media. The method of claim 12,
The computer-readable recording medium which recorded the design program of the photomask further characterized by the process of performing the verification of the part which did not satisfy the design rule in the design pattern data after correction | amendment. The method of claim 12,
A computer-readable recording medium having recorded thereon a program for designing a photomask, wherein the process conversion difference correction is executed. 20. The method of claim 19,
A computer recording a design program of a photomask, characterized in that the optical proximity effect correction is executed on the basis of a previously obtained correlation between at least one of the surrounding length and the pattern area ratio of the pattern and the value of the optical proximity effect correction; Readable recording medium.

Images